360 Degree Assessment & Certification - Q1 2021 - MRG Effitas
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
360 Degree Assessment & Certification Q1 2021 Copyright © 2021 MRG Effitas Ltd. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 1 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
MRG Effitas Ltd. Efficacy Assessment & Assurance Contents Contents ............................................................................................................................................................ 2 Introduction ....................................................................................................................................................... 3 Executive Summary........................................................................................................................................... 4 360 Degree Assessment Certification ......................................................................................................... 5 The Purpose of this Report ........................................................................................................................... 6 Tests Employed ................................................................................................................................................. 7 In the Wild 360 / Full Spectrum Test ............................................................................................................ 7 PUA / Adware Test ........................................................................................................................................ 7 Exploit / Fileless Test .................................................................................................................................... 8 Botnet Test .................................................................................................................................................... 8 Simulator Test ............................................................................................................................................... 8 False Positive Test ........................................................................................................................................ 8 Performance Test.......................................................................................................................................... 8 Security Applications Tested .......................................................................................................................... 10 Malware sample types used ........................................................................................................................... 10 Test Results ..................................................................................................................................................... 11 MRG Effitas is a world-leading, independent IT security efficacy testing In the Wild 360 / Full Spectrum test results .............................................................................................. 11 & assurance company. We are trusted 360 Exploit Degree ...................................................................................................................................... 14 by antimalware vendors across the 360 Exploit Degree Certification ................................................................................................................. 15 world. 360 Online Banking Degree......................................................................................................................... 16 360 Online Banking Degree Certification ................................................................................................... 18 TEL: +44 (0)20 3239 9289 Understanding the Grade of Pass .................................................................................................................. 22 Appendix 1 ....................................................................................................................................................... 24 EMAIL: Appendix 2 ....................................................................................................................................................... 37 contact@mrg-effitas.com TWITTER: @mrgeffitas MRG Effitas 360 Assessment & Certification Programme – Q1 2021 2 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Introduction MRG Effitas is a world leader independent IT research company having a core focus on AV efficacy assessments both in the traditional “Real World” malware detection capabilities and in the financial fraud prevention area. The methodology employed in this test maps closely to Real World practice representing the valid threads endangering anyone using the Windows operating system. This evaluation is aimed to help users choosing the most suitable security application. This Programme is called “360 Assessment & Certification”, since it tests the capabilities of the participating security applications with a full spectrum of attack vectors. In the 360 Assessment, trojans, backdoors, spyware, financial malware, ransomware and “other” malicious applications are all used. Alongside the traditional In-The-Wild (ITW) file-based attacks, our evaluation also contains scenarios where fileless cases and exploitation techniques, live botnets and financial malware simulators are also applied. Besides the malicious attacks, in order to evaluate the practical accuracy of AV products, they were exposed to potentially unwanted applications (PUA or Greyware) and clean files (FP) as well. Additionally, besides security capabilities tests, our assessment measured the footprint each security software on a computer’s performance. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 3 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Executive Summary This Certification Programme is designed to serve as a reflection of product Out of ten tested security products, the following eight managed to meet efficacy based on what we have previously termed “metrics that matter”. the specification to attain our Q1 2021 360 Degree Certification. • Avast Business Antivirus Based on decades of experience in IT security, our previous tests, and being one of the world’s largest supplier of early-life malicious binaries and URLs, • Bitdefender Endpoint Security we know that all endpoints can and will be infected, regardless of the • F-Secure Computer Protection Premium security solutions employed. The question is not ‘if’, but ‘when’ a malicious • Malwarebytes Endpoint Protection binary hits the system. • Microsoft Windows Defender • Sophos Intercept X A security product’s ability to block initial infection (although critical in most • Symantec Endpoint Protection cases) is not the only metric that matters. Measuring the time taken to detect • Trend Micro Security malicious files or actions, is another metric that can also be crucial in evaluation. An additional key factor is the point in time when the fact of the Out of ten tested security products, the following five managed to meet the infection and any associated malicious behaviour are detected. specification to attain our Q1 2021 360 Exploit Degree Certification. • Bitdefender Endpoint Security When conducting these tests, we try to simulate normal user behaviour. We are aware that a “Real World” test cannot be conducted by a team of • ESET Endpoint Security professionals inside a lab, because we understand how certain types of • F-Secure Computer Protection Premium malware work, how campaigns of organised malware attacks are conducted, • Malwarebytes Endpoint Protection and how such attacks could be prevented. Simulating normal user behaviour • Sophos Intercept X means that we pay special attention to all alerts given by security applications. A pass was given only when alerts were straightforward, and Out of ten tested security products, the following three managed to meet clearly suggested that the malicious action should be blocked. the specification to attain our Q1 2021 360 Online Banking Degree Certification. With these in mind, it is very important to note that the best choice for an • Bitdefender Endpoint Security average user is to keep things as simple as possible and not to overwhelm • ESET Endpoint Security the non-tech savvy with cryptic pop-ups, alerts or questions. • Symantec Endpoint Protection MRG Effitas 360 Assessment & Certification Programme – Q1 2021 4 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
360 Degree Assessment Certification In order to attain a quarterly MRG Effitas 360 Degree Level 1 certification, a security application must completely protect the system from initial infection either by automatically blocking every ITW sample, or by blocking them based on their behaviour, prior to any malicious actions and the product must pass the Live Botnet test. PUA, FP, Exploit/Fileless, Financial Malware Simulator, and performance tests are not part of the certification. Level 2 certification is given if the application blocks or detects any initially missed malware in at least 98% of all cases on the 24-hour retest, while the initially missed test cases are less than 10%. If a ransomware/wiper successfully runs and the files are not available anymore, Level 2 certification is lost. Under the MRG Effitas 360 Degree Assessment & Certification, the following products were certified for Q1 2021. Certified (Level 1) • Bitdefender Endpoint Security Certified (Level 2) • Avast Business Antivirus • F-Secure Computer Protection Premium • Malwarebytes Endpoint Protection • Microsoft Windows Defender • Sophos Intercept X • Symantec Endpoint Protection • Trend Micro Security MRG Effitas 360 Assessment & Certification Programme – Q1 2021 5 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
The Purpose of this Report Since its inception in 2009, MRG Effitas has strived to To make testing more akin to real world scenarios, no manual scanning was conducted. Instead, the system was retested exactly 24 hours after the differentiate itself from traditional testing houses by having system was compromised, thereby giving security applications the its primary focus on providing “efficacy assessments” and opportunity to detect infections on restart. not just performing “tests”. As we have stated in our previous test reports, most malware has one primary objective, and that is to make money for the cybercriminals, thus Traditionally, testing of security software has been aimed at measuring a making malware creation a lucrative business with its own unique economic product’s ability to detect malware. Testing has evolved rapidly over the last models and traits1. couple of years, as most labs, under the direction of AMTSO (of which MRG Effitas is a member) has been striving to conduct “Real World” testing, based Measuring initial detection rates and the time taken to detect active malware on standardised guidelines. More information about the compliance status is important, particularly in today’s threat landscape with the mix of malware of this test can be found on the AMTSO website. that is prevalent. Since, the longer a cybercriminal can run their malware on a system, the greater the opportunity is for them to be able to capture private https://www.amtso.org/amtso-ls1-tp035 user information, including banking logins and social media credentials, etc., Although there is no absolute definition of this kind of testing, loosely or to encrypt user data. speaking, it involves the introduction of malware to an endpoint through a For these types of malware, initial detection is of the utmost importance, realistic entry point, such as downloading the sample using a browser or since the vast majority of security solutions will be unable to remediate the getting it from a USB memory stick. Real world testing mostly involves problem of an encrypted system. “dynamic testing” (i.e. the malware is executed and then the ability of the In providing these quarterly certifications, the MRG Effitas 360 Degree security product to block the malware is measured). Assessment & Certification Programme is the de facto standard by which Whilst these types of tests are useful, yielding valid and meaningful data, security vendors, financial institutions and other corporations can attain the MRG Effitas wanted to merge standalone tests and also go the extra mile by most rigorous and accurate determination of a product’s efficacy against the measuring the time security products take to detect infections and remediate full spectrum of malware that is prevalent during the period. the endpoint. 1 For instance, in many ransomware campaigns, the criminals actually operate a 0-24 full blown customer help desk to help victims with buying BitCoin, installing the ToR Browser etc., with a better „user experience” than traditional help desk services. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 6 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Tests Employed In this assessment (Q1 2021), we ran the following tests. In the Wild 360 / Full Spectrum Test PUA / Adware Test Most of the malicious URLs used in this test were compromised legitimate The PUA samples used in this test are deceptive, or potentially unwanted websites, serving malware. We believe that such URLs pose the greatest applications (PUA), that are not malicious, but are generally considered danger to users, as this is the place where they least expect to get infected, unsuitable for most home or business networks. They usually contain and any URL based protection fails on them. Some URLs originate from our adware, installs toolbars or have other unclear objectives. They may also honeypots, or in case of ransomware and financial malware in particular, we contribute to consuming computing resources or network bandwidth. PUAs used URLs from newly discovered distribution sites. can be deceptive, harmful, hoax, show aggressive popups and misleading or scaring the user. They may provide some unconventional ways of Malware delivered by URLs used in this test can be considered as zero-day in uninstalling the application, maybe retain some of their components on the the true meaning of the phrase. This posed a significant challenge to the device without the user's consent. We mainly use a filtered version of participant products. AppEsteem’s feed, as they developed deceptor requirements as part of a cross-industry effort of many of the world’s leading security companies and ~10% of the threats used in this test were introduced to the system via represent a minimum bar that all apps and services must meet to avoid internal webmail sites. We have witnessed many SMBs being infected being titled deceptive. through internal webmails and lack of spam filtering. Downloading malware attachments from internal webmail sites bypass the URL blocking features of AppEsteem, as a member of the AMTSO group is dedicated to help the products, and this happens in-the-wild. protecting consumers from harassing and objectionable material, and to enable security companies to restrict access to such actions. MRG Effitas, as During the In the Wild 360 / Full Spectrum test, 360 live ITW samples were a member of the AMTSO group, is also dedicated to protecting these used. The stimulus load comprised the following: 34 trojans, 45 backdoors, thoughts. 42 financial malware samples, 53 ransomware samples, 41 spyware, 49 malicious documents, 22 malicious emails, 74 malicious script files. In the PUA/Adware part we tested the products against 10 PUAs. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 7 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Exploit / Fileless Test Simulator Test - Magecart The main purpose of this test is to see how security products protect against The Financial malware simulator used in 2021Q1 was the Magecart credit a specific exploitation technique. In order to measure this, we developed test card-skimming attack. Magecart is the name of the collection of groups who cases that simulate the corresponding exploit and post-exploitation are targeting some eCommerce sites and ticketing companies such as techniques only. Ticketmaster, British Airways, Newegg, Infowars etc. They used small JavaScript codes on the online store’s checkout pages. These scripts pulled Drive-by download exploits are the biggest threats for an enterprise personal and/or credit card data and sent it to the attackers’ servers. environment, since no user interaction is needed to start the chain of infection on a victim machine. Outdated browsers and Office environments are widespread in enterprise environments, due to compatibility issues or the lack of proper patch management process. False Positive Test Perfect blocking of malicious content is only part of the story from a We were testing the products’ abilities to avoid any exposure to adversaries, practical point of view for any decent AV product. In many cases all malware to interrupt malicious payload delivery before performing malicious actions. blocking is a result of a very aggressive filter which can block non-malicious We focus explicitly on each product’s ability to mitigate each attack legitimate applications as well prohibiting everyday work by blocking technique. The results are not intended to evaluate the complete efficacy of legitimate, perhaps newly developed in-house software. the products, but rather the products’ anti-exploit and anti-post-exploit features in isolation. In order to test this feature, we tested the security applications against completely clean, recently created applications. During this test we used 8 different exploitation techniques. The detailed False positive assessment consisted of 1000 clean and legitimate description can be found in the ‘Appendix’. application samples. The selection has been focused on applications, frequently found in enterprise environments (drivers, media editors, developer tools, etc.) Botnet Test A python based BYOB (Build Your Own Botnet) inspired tool was used as Botnet test. Its behavior is parallel to any in-the-wild botnet. The main built-in Performance Test feature is designed to steal credentials with its two major components, a A security product’s usefulness does not depend on protection level solely, CnC Server, and a downloaded Portable Executable file. but also on its resource footprint and its effect of the overall operating system performance. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 8 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
In order to assess the products’ influence on the operating system, we tested several performance factors on a physical machine and combined the results, based on a scoring approach. Detailed information can be found in the ‘Appendix’. In every test case, (except for the performance test) our testing environment supports the execution of VM-aware malware, this is the reason why we were able to use more sophisticated threats which normally would not run on Virtual Machines. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 9 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Security Applications Tested Malware sample types used to conduct the tests • Avast Business Antivirus 20.10.2625 • Avira Antivirus Pro 15.0.2101.2069 ITW samples distribution • Bitdefender Endpoint Security 6.6.25.362 • ESET Endpoint Security 7.3.2036.0 6,1% • F-Secure Computer Protection Premium 21.1 (4.12.13.68.0/PSB) 11,7% • Malwarebytes Endpoint Protection 1.2.0.848 • Microsoft Windows Defender 4.18.1911.3 13,6% Financial malware • Sophos Intercept X 2.0.20 • Symantec Endpoint Protection 14.3.3384.1000 Ransomware • Trend Micro Security 6.7.1524/14.2.1282 14,7% Backdoor Spyware Malicious script Malicious document 20,6% 12,5% Malicious email 11,4% MRG Effitas 360 Assessment & Certification Programme – Q1 2021 10 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Test Results The tables below show the results of testing under the MRG Effitas 360 Q1 2021 Assessment Programme. In the Wild 360 / Full Spectrum test results The table below shows the detection rates of the security products for 360 ITW samples. This table is sorted by smallest number of missed samples. In the Wild 360 / Full Spectrum test results 100% 90% 80% Malwarebytes Trend Micro Bitdefender Symantec Microsoft 70% F-Secure Sophos Avast ESET Avira 60% Bitdefender Malwarebytes Trend Micro Avast Symantec Sophos F-Secure Microsoft ESET Avira Miss 0,00% 0,00% 0,28% 0,28% 0,56% 0,56% 0,83% 0,83% 1,11% 11,11% Blocked in 24h 0,00% 0,00% 0,00% 0,00% 0,00% 0,56% 0,00% 1,94% 1,94% 1,11% Behaviour block 5,00% 28,89% 0,56% 2,22% 3,06% 0,56% 3,89% 0,00% 2,78% 15,28% Auto block 95,00% 71,11% 99,17% 97,50% 96,39% 98,33% 95,28% 97,22% 94,17% 72,50% MRG Effitas 360 Assessment & Certification Programme – Q1 2021 11 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Ransomware samples test results The table below shows the detection rates of the security products for 53 ransomware samples. This table is sorted by smallest number of missed samples. Ransomware samples test results 100% 90% 80% Malwarebytes Trend Micro Bitdefender 70% Symantec Microsoft F-Secure Sophos Avast ESET Avira 60% Bitdefender F-Secure Microsoft Sophos Symantec Trend Micro Avast Malwarebytes ESET Avira Miss 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 3,77% Blocked in 24h 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 7,55% 0,00% Behaviour block 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 1,89% 3,77% 0,00% 20,75% Auto block 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 98,11% 96,23% 92,45% 75,47% MRG Effitas 360 Assessment & Certification Programme – Q1 2021 12 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
PUA/adware samples test results The table below shows the detection rates of the security products for 10 PUA/Adware samples. This table is sorted by smallest number of missed samples. PUA/Adware samples test results 100% 90% 80% 70% 60% 50% 40% Malwarebytes Trend Micro Bitdefender 30% Symantec Microsoft F-Secure 20% Sophos Avast ESET Avira 10% 0% Avast Avira Bitdefender ESET Malwarebytes Microsoft Sophos Symantec F-Secure Trend Micro Miss 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 30,00% 60,00% Blocked in 24h 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% Behaviour block 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% Auto block 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 70,00% 40,00% MRG Effitas 360 Assessment & Certification Programme – Q1 2021 13 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
360 Exploit Degree The table below shows the initial detection rates of the security products for 8 exploit / fileless test. This table is sorted by smallest number of missed attack vectors. Exploit/Fileless samples test results 100% 90% 80% 70% 60% 50% 40% Malwarebytes Trend Micro Bitdefender 30% Symantec Microsoft F-Secure Sophos 20% Avast ESET Avira 10% 0% ESET Sophos F-Secure Malwarebytes Bitdefender Avast Avira Microsoft Symantec Trend Micro Miss 0,00% 0,00% 0,00% 0,00% 0,00% 12,50% 12,50% 12,50% 12,50% 25,00% Signature block 0,00% 0,00% 12,50% 37,50% 62,50% 0,00% 0,00% 0,00% 0,00% 0,00% Behaviour block 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 12,50% Auto block 100,00% 100,00% 87,50% 62,50% 37,50% 87,50% 87,50% 87,50% 87,50% 62,50% MRG Effitas 360 Assessment & Certification Programme – Q1 2021 14 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
360 Exploit Degree Certification In order to attain a quarterly MRG Effitas 360 Exploit Degree certification award, a security application must entirely protect the system from initial infection (autoblock, signature block, or behaviour block). Certified • Bitdefender Endpoint Security • ESET Endpoint Security • F-Secure Computer Protection Premium • Malwarebytes Endpoint Protection • Sophos Intercept X MRG Effitas 360 Assessment & Certification Programme – Q1 2021 15 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
360 Online Banking Degree Financial malware samples test results The table below shows the detection rates of the security products for 42 financial malware samples. This table is sorted by smallest number of missed samples. Financial malware samples test results 100% 90% 80% Malwarebytes Trend Micro Bitdefender Symantec Microsoft 70% F-Secure Sophos Avast ESET Avira 60% Bitdefender ESET F-Secure Malwarebytes Microsoft Sophos Symantec Trend Micro Avast Avira Miss 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 7,14% Blocked in 24h 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% Behaviour block 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 2,38% 14,29% Auto block 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 97,62% 78,57% MRG Effitas 360 Assessment & Certification Programme – Q1 2021 16 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Real Botnet test results Financial malware simulator test results The table below shows the results of live financial malware test. The table shows the results of financial malware simulator test. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 17 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
360 Online Banking Degree Certification In order to attain a quarterly MRG Effitas 360 Online Banking Degree certification award, a security application must entirely protect the system from initial In-the-wild financial malware infection (autoblock or behaviour block) and the product must pass the Botnet and Financial malware simulator tests during the quarter. Under the MRG Effitas 360 Degree Assessment & Certification, the following products were certified for Q1 2021. Certified • Bitdefender Endpoint Security • ESET Endpoint Security • Symantec Endpoint Protection MRG Effitas 360 Assessment & Certification Programme – Q1 2021 18 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
False positive samples test results The table below shows the initial detection rates of the security products for 1000 false positive samples. This table is sorted by smallest number of false positive sample blocks. False positive samples test results 100% Malwarebytes Trend Micro Bitdefender Symantec Microsoft F-Secure Sophos Avast ESET Avira 90% ESET F-Secure Microsoft Sophos Symantec Trend Micro Avast Malwarebytes Avira Bitdefender False block 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,09% 0,18% 0,27% 0,00% Detected 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 0,00% 3,55% Allowed to run 100,00% 100,00% 100,00% 100,00% 100,00% 100,00% 99,91% 99,82% 99,73% 96,45% MRG Effitas 360 Assessment & Certification Programme – Q1 2021 19 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Performance test results This table is sorted from highest to lowest score where the highest score denotes the lowest impact on the system. Performance test results Malwarebytes Trend Micro Bitdefender Symantec Microsoft F-Secure Sophos Avast ESET Avira ESET Microsoft F-Secure Avast Malwarebytes Symantec Bitdefender Trend Micro Sophos Avira Chart ordering 9,74 8,68 7,63 7,58 7,53 7,05 6,95 6,84 6,68 6,32 Scoring details can be found in the ‘Appendix’. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 20 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Detailed results of the performance test The table below shows the detailed results of the performance test of the security products. This table is sorted alphabetically. Windows 10 Avast Avira Bitdefender ESET F-Secure Malwarebytes Microsoft Sophos Symantec Trend Micro Base Bootup time (s) 30,4 39,4 40,3 39,4 35,1 33,3 43,2 30,8 49,0 37,0 44,3 Security software size on disk (Mb) n/a 1384,5 968,8 977,7 738,5 516,8 416,9 349,8 2196,4 3381,6 739,1 Browser Operations (s) Website Open 2,5 3,5 3,6 2,9 3,6 2,9 3,3 2,3 4,0 7,7 4,6 File Download 11,4 15,4 11,9 12,1 11,3 10,9 14,8 11,0 11,8 13,8 12,6 File Operations (s) File Copy 2,0 2,1 2,4 2,0 1,9 1,9 1,8 2,3 2,0 1,9 2,4 File Compression 38,5 37,9 38,1 38,1 37,7 37,9 47,1 38,1 39,5 38,7 52,4 Archive Extraction 6,5 6,0 7,1 7,1 5,9 10,1 6,9 9,9 18,0 10,2 28,8 Office File Opening (s) Excel 6,2 6,8 7,3 7,5 6,5 6,9 6,9 6,9 7,4 6,7 8,7 Word 1,1 2,7 3,6 3,1 2,6 3,0 3,2 2,7 3,4 3,5 5,9 Security software update Time (s) n/a 36,0 51,3 289,0 13,7 17,3 n/a 34,0 108,0 18,3 24,3 CPU usage (%) n/a 36,4 34,5 29,8 40,7 50,7 n/a 20,6 31,6 45,6 26,4 Memory usage (Mb) n/a 411,0 521,8 1182,9 0,0 467,6 n/a 97,3 773,2 238,8 383,0 Physical disk usage (%) n/a 28,6 11,0 13,5 155,2 38,6 n/a 10,0 9,3 12,0 5,5 Network interface usage (B/s) n/a 43703,6 150207,9 58517,0 51890,0 177020,8 n/a 129921,4 7449,7 338701,6 134690,8 Security software scanning - C:\ Time (s) n/a 359,7 473,3 172,0 147,7 501,0 799,7 691,0 610,3 445,3 24,3 CPU usage (%) n/a 21,4 29,4 43,0 24,9 76,0 37,2 90,9 27,4 22,1 26,4 Memory usage (Mb) n/a 866,4 844,6 1876,2 135,6 600,2 742,3 757,5 1475,8 449,4 383,0 Physical disk usage (%) n/a 27,3 21,3 11,9 10,2 29,9 8,0 16,4 4,8 12,8 5,5 Network interface usage (B/s) n/a 2075,7 107259,8 638,3 1436,2 2140,2 367,0 600,4 608,3 2521,6 134690,8 MRG Effitas 360 Assessment & Certification Programme – Q1 2021 21 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Understanding the Grade of Pass 360 Degree Level 1 certified 360 Degree Level 2 certified 360 Degree Not certified All threats detected on first exposure or via At least 98% of the threats detected and Security product failed to detect at least 98% of behaviour protection. neutralised / system remediated before or on the infections and remediate the system during the first rescan while the initially missed test the test procedure, or at least one ransomware cases are less than 10% and no ransomware was missed. was missed on first exposure. • Bitdefender Endpoint Security • Avast Business Antivirus • Avira Antivirus Pro • F-Secure Computer Protection Premium • ESET Endpoint Security • Malwarebytes Endpoint Protection • Microsoft Windows Defender • Sophos Intercept X • Symantec Endpoint Protection • Trend Micro Security MRG Effitas 360 Assessment & Certification Programme – Q1 2021 22 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
360 Exploit Degree Certified 360 Exploit Degree Not Certified The application entirely protected the system from initial infection. The application failed to protect the system from initial infection. • Bitdefender Endpoint Security • Avast Business Antivirus • ESET Endpoint Security • Avira Antivirus Pro • F-Secure Computer Protection Premium • Microsoft Windows Defender • Malwarebytes Endpoint Protection • Symantec Endpoint Protection • Sophos Intercept X • Trend Micro Security 360 Online Banking Degree Certified 360 Online Banking Degree Not Certified The application entirely protected the system from initial In-the-wild The application failed to protect the system from initial In-the-wild financial financial malware infection and passed the Botnet and Financial malware infection, or it has not passed the Botnet or Financial malware malware simulator test. simulator test. • Bitdefender Endpoint Security • Avast Business Antivirus • ESET Endpoint Security • Avira Antivirus Pro • Symantec Endpoint Protection • F-Secure Computer Protection Premium • Malwarebytes Endpoint Protection • Microsoft Windows Defender • Sophos Intercept X • Trend Micro Security MRG Effitas 360 Assessment & Certification Programme – Q1 2021 23 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Appendix 1 Methodology used in the “In the Wild 360 / Full Spectrum” test 1. Windows 10 Enterprise 64-bit operating system is installed on a hardened virtual machine, all updates are applied, and third-party applications installed and updated. 2. An image of the operating system is created. 3. A clone of the imaged systems is made for each of the security applications used in the test. 4. An individual security application is installed using default settings on each of the systems created in (3) and then, where applicable, updated. If the vendor provided a non-default setting, this setting is checked whether it is realistic. If yes, the changes are documented, applied, and added to the appendix section of the report. 5. A clone of the system as at the end of (4) is created. 6. Downloading a single binary executable (or document, script, etc.) from its native URL using Chrome to the Downloads folder and then executing the binary in the clean, unprotected system. If the sample works, the sample is saved in a replay proxy to provide the same binary throughout the test. Live URL test is conducted by the following procedure. 6.1. The sample is selected for the test and tested in the systems where a security product is installed. 6.2. The test case is retested 24 hours after the initial test if the security application failed to block the malicious binary. Spam e-mail attachment test is conducted by the following procedure. 6.3. Microsoft Office Outlook client downloading a single email from its server to the victim system created in (4). 6.4. Opening the e-mail, saving the attachment to the Downloads folder and then executing the binary. 6.5. The test case is retested 24 hours after the initial test if the security application failed to block the malicious binary. • The test case is marked as “Blocked” by either the security application blocks the URL where the malicious binary was located. Or the security application blocks the malicious binary whilst it was being downloaded to the desktop. • The test case is marked as “Behaviour Blocked” if the security application blocks the malicious binary when it is executed and either automatically blocks it or postpones its execution and warns the user that the file is malicious and awaiting user input. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 24 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
• The test case is marked as “Detected” if the security application detects the threat and sends an alert to the central console or notifies the user, but the sample is allowed to run. • The test case is marked as “Blocked in 24h” if the security application fails to block or behaviour block the malicious sample but blocks it during the retest. • The test case is marked as “Missed” if the security application fails to block or behaviour block the malicious sample during both tests. 7. Tests are conducted with all systems having internet access. 8. As no user-initiated scans is involved in this test, applications rely on various technologies to detect, block and remediate threats. Some of these technologies are URL blacklisting, reputation, signature, machine learning, heuristics, behaviour etc. Methodology used in the “In-The-Wild PUA/Adware” test 1. Windows 10 Enterprise 64-bit operating system is installed on a hardened virtual machine, all updates are applied, and third-party applications installed and updated. 2. An image of the operating system is created. 3. A clone of the imaged systems is made for each of the security applications used in the test. 4. An individual security application is installed using default settings on each of the systems created in (3) and then, where applicable, updated. If the vendor provided a non-default setting, this setting is checked whether it is realistic. If yes, the changes are documented, applied, and added to the appendix section of the report. 5. A clone of the system as at the end of (4) is created. 6. Downloading a single binary executable (or document, script, etc.) from its native URL using Chrome to the Downloads folder and then executing the binary in the clean, unprotected system. If the sample works, the sample is saved in a replay proxy to provide the same binary throughout the test. 7. The sample is selected for the test and tested in the systems where a security product is installed. 8. The test case is retested 24 hours after the initial test if the security application failed to block the malicious binary. • The test case is marked as “Blocked” by either the security application blocks the URL where the malicious binary was located. Or the security application blocks the malicious binary whilst it was being downloaded to the desktop. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 25 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
• The test case is marked as “Behaviour Blocked” if the security application blocks the malicious binary when it is executed and either automatically blocks it or postpones its execution and warns the user that the file is malicious and awaiting user input. • The test case is marked as “Detected” if the security application detects the threat and sends an alert to the central console or notifies the user, but the sample is allowed to run. • The test case is marked as “Blocked in 24h” if the security application fails to block or behaviour block the malicious sample but blocks it during the retest. • The test case is marked as “Missed” if the security application fails to block or behaviour block the malicious sample during both tests. 9. Tests are conducted with all systems having internet access. As no user-initiated scans is involved in this test, applications rely on various technologies to detect, block and remediate threats. Some of these technologies are URL blacklisting, reputation, signature, machine learning, heuristics, behaviour etc. Methodology used in the false positive test 1. Windows 10 Enterprise 64-bit operating system is installed on a hardened virtual machine, all updates are applied, and third-party applications installed and updated. 2. An image of the operating system is created. 3. A clone of the imaged systems is made for each of the security applications used in the test. 4. An individual security application is installed using default settings on each of the systems created in (3) and then, where applicable, updated. If the vendor provided a non-default setting, this setting is checked whether it is realistic. If yes, the changes are documented, applied, and added to the appendix section of the report. 5. A clone of the system as at the end of (4) is created. 6. Introducing the binary executables (or documents, scripts, etc.) to the clean, unprotected system via disk image or network share. If the sample works, the sample is saved to a different disk image or network share. False Positive test is conducted by the following procedure. 6.1. Scanning the binary executables (or documents, scripts, etc.) on the disk image or on the network share. 6.2. Executing the test samples. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 26 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
6.3. The sample is retested 24 hours after the initial test if the security application failed to permit the harmless file. • The test case is marked as “False block” if the security application falsely identifies and blocks the binary at any stage during the test and retest. • The test case is marked as “Detected” if the security application falsely identifies and the binary at any stage during the test and retest but allows it to run. • The test case is marked as “Allowed to run in 24h” if the security application falsely identifies and blocks the binary at any stage during the test but allows it to run upon the retest. • The test case is marked as “Allowed to run” if the security application correctly identifies the binary as harmless and allows it to run. 7. Tests are conducted with all systems having internet access. Methodology used in the exploit/fileless test 1. Windows 10 Enterprise 64-bit operating system is installed on a hardened virtual machine, all updates are applied, and third-party applications installed and updated. 2. An image of the operating system is created. 3. A clone of the imaged systems is made for each of the security applications used in the test. 4. An individual security application is installed using default settings on each of the systems created in (3) and then, where applicable, updated. If the vendor provided a non-default setting, this setting is checked whether it is realistic. If yes, the changes are documented, applied, and added in the report in an appendix. 5. A clone of the system as at the end of (4) is created. Exploit / Fileless test is conducted by the following procedure. 6. Our payloads use an exploit for the one of an installed vulnerable application. In order to simulate a realistic attack scenario, a payload is constructed to include at least one of the common CnC frameworks. 7. The opening stage of the exploit is introduced to the system and we monitor if the vulnerable application starts the initial stage payload, the exploit is being executed and if a session is established to our CnC server. 8. After navigating to the exploit site, the system is supervised if there are any new processes, loaded DLLs or CnC traffic emerge. If the exploitation is successful, the following actions are executed. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 27 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
8.1. Upload a file to the victim. 8.2. Download a file from the victim. 8.3. Create a process remotely. 8.4. Read the contents of a file on the victim. 9. When user interaction is needed from the endpoint protection (e.g. site visit not recommended, etc.) the default action is chosen. When user interaction is needed from the operating system, we chose the run/allow options. 10. Throughout the test, the Process Monitor from the Sysinternals Suite and Wireshark are running (both installed to non-default directories and modified not to be detected by default anti-debugging tools). • The test case is marked as “Signature Block” if the security application blocks the URL (infected URL, exploit kit URL, redirection URL, malware URL) by the URL database (local or cloud). • The test case is marked as “Blocked” if the security application blocks the page containing a malicious HTML code, JavaScript (redirects, iframes, obfuscated JavaScript, etc.) or Flash files. Or if the security application blocks the downloaded payload by analysing the malware before it can be started. (reputation-based block or heuristic based block). • The test case is marked as “Behaviour Blocked” if the security application blocks the downloaded payload after it has been started. • The test case is marked as “Detected” if the security application detects the threat and sends an alert to the central console or notifies the user, but the attack is allowed to run. • The test case is marked as “Missed” if the security application fails to detect, block or behaviour block the attack and the it can be carried out. 11. Tests are conducted with all systems having internet access. 12. As no user-initiated scans is involved in this test, applications rely on various technologies to detect, block and remediate threats. Some of these technologies are URL blacklisting, reputation, signature, machine learning, heuristics, behaviour etc. Detailed description of the Exploit / Fileless cases. Test case 001 Koadic / WMIC Koadic is a framework using VBScript stagers for increased stealth and limited footprint. In this test case, a Koadic connectback payload is instantiated using a wmic command. MRG Effitas 360 Assessment & Certification Programme – Q1 2021 28 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
In case the exploitation was successful, as a proof of that working session has been establised, the following actions were carried out through the connection. - A directory list is queried - A file is uploaded to the victim - A file is downloaded - A shell command is executed The test case is flagged as MISSED if exploitation was successful and test machine had been successfully controlled via the new session. References: https://github.com/zerosum0x0/koadic Test case 002 Koadic / MSHTA Koadic is a framework using VBScript stagers for increased stealth and limited footprint. In this test case, a Koadic connectback payload is instantiated using a malicious Windows help .hta document. In case the exploitation was successful, as a proof of that working session has been established, the following actions were carried out through the connection. - A directory list is queried - A file is uploaded to the victim - A file is downloaded - A shell command is executed The test case is flagged as MISSED if exploitation was successful and test machine had been successfully controlled via the new session. References: https://github.com/zerosum0x0/koadic Test case 003 Koadic / regsvr32 Koadic is a framework using VBScript stagers for increased stealth and limited footprint. In this test case, a Koadic connectback payload is instantiated using a regsvr32 remote object load call. In case the exploitation was successful, as a proof of that working session has been established, the following actions were carried out through the connection. - A directory list is queried MRG Effitas 360 Assessment & Certification Programme – Q1 2021 29 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
- A file is uploaded to the victim - A file is downloaded - A shell command is executed The test case is flagged as MISSED if exploitation was successful and test machine had been successfully controlled via the new session. References: https://github.com/zerosum0x0/koadic Test case 004 Octopus / WMIC In this test case, we use the Octopus framework which is a framework using VBScript stagers for increased stealth and limited footprint. In this test case, a Powershell connectback payload is instantiated. In case the exploitation was successful, as a proof of a working session, the following steps were taken. - A directory list is queried - A file has been downloaded - A file has been uploaded - A shell command is executed The test case is flagged as MISSED if exploitation was successful and the test machine had been successfully controlled via the new session. References: https://github.com/mhaskar/Octopus Test case 005 Octopus / MSHTA In this test case, we use the Octopus framework which is a framework using VBScript stagers for increased stealth and limited footprint. In this test case, a Powershell connectback payload is instantiated. In case the exploitation was successful, as a proof of a working session, the following steps were taken. - A directory list is queried - A file has been downloaded - A file has been uploaded - A shell command is executed The test case is flagged as MISSED if exploitation was successful and the test machine had been successfully controlled via the new session. References: https://github.com/mhaskar/Octopus MRG Effitas 360 Assessment & Certification Programme – Q1 2021 30 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Test case 006 Foxit reader Use After Free + Empire In this test case, we use the Foxit Reader v9.0.1.1049 exploit (foxit_reader_uaf) to start the exploit chain. After successfully exploiting the vulnerability an Empire (PowerShell) stager is executed. In case the exploitation was successful, as a proof of a working session, the following steps were taken. - A screenshot has been made - A file has been downloaded - A file has been uploaded The test case is flagged as MISSED if exploitation was successful and the test machine had been successfully controlled via the new session. Exploited application: Foxit Reader v9.0.1.1049 OS version: Windows 7 CVE: CVE-2018-9948 CVE-2018-9958 The exploit Foxit Reader v9.0.1.1049 and earlier are affected by use-after-free and uninitialized memory vulnerabilities that can be used to gain code execution. This module uses Uint32Array uninitialized memory and text annotation use-after-free vulnerabilities to call WinExec with a share file path to download and execute the specified exe. The module has been tested against Foxit Reader v9.0.1.1049 running on Windows 7 x64 and Windows 10 Pro x64 Build 17134. Windows 10 Enterprise needs to have insecure logons enabled for the exploit to work as expected. References: https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2018-9948 https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2018-9958 https://www.rapid7.com/db/modules/exploit/windows/fileformat/foxit_reader_uaf https://www.powershellempire.com/ https://github.com/EmpireProject/Empire. Test case 007 Firefox version 31.0 exploit with Empire In this test case, we target Firefox 31.0 with an exploit (CVE-2014-8636, CVE-2015-0802) starting the exploit chain. After successfully exploiting the vulnerability an Empire (PowerShell) stager is executed. In case the exploitation was successful, as a proof of a working session, the following steps were taken. - A screenshot has been made - A file has been downloaded MRG Effitas 360 Assessment & Certification Programme – Q1 2021 31 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
- A file has been uploaded The test case is flagged as MISSED if exploitation was successful and the test machine had been successfully controlled via the new session. The exploit This exploit gains remote code execution on Firefox 31-34 by abusing a bug in the XPConnect component and gaining a reference to the privileged chrome:// window. This exploit requires the user to click anywhere on the page to trigger the vulnerability. CVE: CVE-2014-8636 CVE-2015-0802 References: https://www.rapid7.com/db/modules/exploit/multi/browser/firefox_proxy_prototype https://www.powershellempire.com/ https://github.com/EmpireProject/Empire Test case 008 MSBuild + Metasploit Meterpreter In this test case, we target MSBuild starting the exploit chain. Assuming that MSBuild.exe is allowed since this tool is part of the Microsoft .NET Framework, we can invoke it to execute a .xml file as a Visual Studio .NET C# Project descriptor. The well-composed file contains a CSharp code which starts a Meterpreter stager. If code execution is not blocked, as a result, a new Meterpreter session back to MRG-Effitas CnC server will be created. In case the exploitation was successful, as a proof of a working session, the following steps are taken. - A screenshot has been made - A file has been downloaded - A file has been uploaded The test case is flagged as MISSED if exploitation was successful and the test machine had been successfully controlled via the new session. References: https://ired.team/offensive-security/code-execution/using-msbuild-to-execute-shellcode-in-c MRG Effitas 360 Assessment & Certification Programme – Q1 2021 32 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
Methodology used in the Real Botnet Test 1. Windows 10 Enterprise 64-bit operating system is installed on a hardened virtual machine, all updates are applied, and third-party applications installed and updated. 2. An image of the operating system is created. 3. A Real botnet dropper is run on the clean, unprotected system, thus simulating a pre-infected state. 4. A clone of the imaged system is made for each of the security applications to be used in the test. 5. An individual security application is installed using default settings on each of the systems created in (4) and then, where applicable, updated. If the vendor provided a non-default setting, this setting is checked whether it is realistic. If yes, the changes are documented, applied, and added in the report in an appendix. 6. A clone of the system as at the end of (5) is created. Real botnet test is conducted by the following procedure. 6.1. Starting a new instance of Firefox (or the Safe Browser) and navigating to a financial website. Where the security application offers a secured or dedicated banking browser, this is used. If the security application is designed to protect Internet Explorer, only that component is tested. 6.2. Text is entered into the Account login page of the financial website using the keyboard or using a virtual keyboard if the application under test provides such functionality, and then the “log in” button is pressed. • The test case is marked as passed – a green checkmark if the security application detects the financial malware when the security application is installed, and a mandatory scan is made. Or the security application detects the real financial malware when it is executed according to the following criteria: o It identifies the real financial malware as being malicious and either automatically blocks it or postpones its execution, warns the user that the file is malicious and awaits user input. o It identifies the real financial malware as suspicious or unknown and gives the option to run in a sandbox or safe restricted mode, which prevents the real financial malware from capturing and sending the logon data to the MRG CnC, whilst giving no alerts or giving informational alerts only. Or The security application intercepts the action of the real financial malware and displays warnings and user action input requests that are clearly different from those displayed in response to legitimate applications. a. The test case is marked as missed – a red cross if the security application fails to detect the real financial malware according to the following criteria: MRG Effitas 360 Assessment & Certification Programme – Q1 2021 33 Copyright © 2021 MRG Effitas Ltd. This article or any part thereof may not be published or reproduced without the consent of the copyright holder
You can also read